PICALM (CALM) - the clathrin-assembly adaptor that is the 5' partner of CALM::AF10

This is a prediction, not a measurement. No one has seen this protein in this shape. The colours say how much the software trusts itself, and that is worth reading before anything else — the orange stretches are places the model is effectively saying this part has no fixed shape.

• A PICALM-directed intervention is predicted to hit erythropoiesis first: Picalm is required for transferrin-receptor endocytosis and erythroid maturation in mice, and knockout animals are anaemic [E6][E7].
• PICALM is not a selective dependency of its own fusion-bearing line in DepMap, so a 'drug the 5' partner' strategy has no dependency signal behind it [E11].
• The PICALM locus is an Alzheimer's risk locus with proposed roles in amyloid clearance [E8]; the CNS consequences of chronic systemic PICALM inhibition are unbounded.
• There is no human PICALM structure to design against [E9], so any structure-guided campaign starts on a rat template.

1. What does PICALM do in the nucleus? UniProt annotates a nuclear pool and PIMREG-dependent nuclear targeting and no functional role has been assigned - yet the fusion's entire contribution from PICALM is a nuclear-export signal.
2. Is there any human PICALM structure? As of 2026-09-20 the PDB has none; every deposit is rat. A human ANTH-domain structure would be a prerequisite for designing against this surface.
3. Does loss of one PICALM allele (the translocated one) contribute anything by itself? No study in this beat's reading separates PICALM haploinsufficiency from gain of the fusion.

1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

   Human PICALM is UniProt Q13492 (PICAL_HUMAN), primary gene name PICALM, gene synonym CALM, 652 aa, annotated to cell membrane, clathrin-coated pit, Golgi apparatus, clathrin-coated vesicle and nucleus.

   UniProtKB Q13492 (PICAL_HUMAN) - Phosphatidylinositol-binding clathrin assembly protein · UniProt Consortium 2026
   ▸What this evidence cannot show

   Fetched live 2026-09-20 and the primary gene name was read back, per brief rules 62-65. UniProt's own N-terminal domain call is 'ENTH 14-145', which conflicts with the Pfam/InterPro ANTH call at 23-283 and with the ANTH naming used throughout the literature; both are recorded in `protein-picalm-domains`.

   ▸Evidence that argues against this (1)

   • A REVIEWED DATABASE RECORD FIXES IDENTITY ON THE DATE IT WAS READ AND NOTHING MORE. It does not establish which isoform is expressed in leukaemic blasts, nor that the wild-type protein's properties survive into the fusion - and for PICALM the fusion retains only part of the protein, so most of what this record describes is on the discarded side. Oncofusion knowledge graph (internal entry) 2026
2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

   PICALM is ENSG00000073921 on chromosome 11 at 85,957,157-86,069,882, minus strand, GRCh38, HGNC:15514.

   Ensembl REST lookup ENSG00000073921 (PICALM), GRCh38 · Ensembl 2026
   ▸What this evidence cannot show

   Fetched live 2026-09-20. The gene is on the MINUS strand, which matters when reasoning about breakpoint orientation in t(10;11)(p13;q14).

   ▸Evidence that argues against this (1)

   • A REVIEWED DATABASE RECORD FIXES IDENTITY ON THE DATE IT WAS READ AND NOTHING MORE. It does not establish which isoform is expressed in leukaemic blasts, nor that the wild-type protein's properties survive into the fusion - and for PICALM the fusion retains only part of the protein, so most of what this record describes is on the discarded side. Oncofusion knowledge graph (internal entry) 2026
3. E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

   CALM localises to endocytic coated pits and the Golgi area, binds clathrin, and its overexpression perturbs clathrin-mediated traffic.

   Clathrin assembly lymphoid myeloid leukemia (CALM) protein: localization in endocytic-coated pits, interactions with clathrin, and the impact of overexpression on clathrin-mediated traffic. · Tebar F, et al. · Mol Biol Cell 1999
   PMID 10436022DOI
   ▸What this evidence cannot show

   Overexpression and immunofluorescence in cultured non-haematopoietic cells. It establishes localisation and clathrin binding, not physiological requirement - that came later from the knockouts [E6][E7].

   ▸Evidence that argues against this (1)

   • The coated-pit localisation result comes from overexpression in non-haematopoietic cells and implies a general endocytic requirement that the genetics does not support: conditional Picalm knockout leaves myeloid and B-lymphoid development intact and confines the requirement to erythroid transferrin-receptor uptake. CALM is a cell-type-restricted adaptor, not a general one. Ishikawa Y et al. 2015
4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

   PICALM-deficient murine embryonic fibroblasts from fit1 mutant mice show increased surface transferrin receptor, decreased intracellular iron, reduced proliferation and sensitivity to iron chelation, all rescued by retroviral PICALM.

   The PICALM protein plays a key role in iron homeostasis and cell proliferation. · Scotland PB, et al. · PLoS One 2012
   PMID 22952941DOI
   ▸What this evidence cannot show

   Murine fibroblasts, not human haematopoietic cells. The paper also shows the C-terminal residues are required for clathrin association - and that C-terminus is exactly what CALM::AF10 keeps.

   ▸Evidence that argues against this (1)

   • WILD-TYPE PICALM ENDOCYTOSIS IN A FIBROBLAST IS NOT FUSION BIOLOGY. fit1 mutant MEFs report what full-length PICALM does in a non-haematopoietic cell; PICALM::MLLT10 discards most of PICALM and relocates the remainder to chromatin. This graph separately records that CALM-AF10 is predominantly CYTOPLASMIC, which makes the trafficking literature relevant to the protein and still not to the leukaemic mechanism. Oncofusion knowledge graph (internal entry) 2026
5. E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

   CALM regulates clathrin-coated-vesicle size and maturation by directly sensing and driving membrane curvature.

   CALM regulates clathrin-coated vesicle size and maturation by directly sensing and driving membrane curvature. · Miller SE, et al. · Dev Cell 2015
   PMID 25898166DOI
   ▸What this evidence cannot show

   Liposome and cell-biological work using largely the isolated ANTH domain; the structures that underpin the curvature model are rat protein [E9].

   ▸Evidence that argues against this (1)

   • MEMBRANE-CURVATURE SENSING IS A PROPERTY OF THE DOMAIN THE FUSION LARGELY DISCARDS, and the assay is reconstituted biophysics rather than leukaemic cells. It is strong for what it measures and carries no weight for the fusion's transcriptional output. Oncofusion knowledge graph (internal entry) 2026
6. E6strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

   In conditional Picalm knockout mice, Picalm is required for transferrin-receptor endocytosis in erythroid cells and for erythroid development, while being dispensable for myeloid and B-lymphoid development; global adult inactivation causes anaemia and a coat-colour change without gross loss of fitness.

   Role of the clathrin adaptor PICALM in normal hematopoiesis and polycythemia vera pathophysiology. · Ishikawa Y, et al. · Haematologica 2015
   PMID 25552701DOI
   ▸What this evidence cannot show

   Mouse genetics. The lineage restriction is the load-bearing part for this beat: a PICALM-directed intervention is predicted to hit erythropoiesis first. The same paper shows Picalm deletion abrogates the disease phenotype in a Jak2 V617F polycythaemia vera model, which is a different disease from this beat's.

   ▸Evidence that argues against this (1)

   • A CONDITIONAL KNOCKOUT MEASURES LOSS OF THE WILD-TYPE GENE, NOT GAIN OF THE FUSION. Picalm is required for transferrin-receptor endocytosis in mice, which bears on haploinsufficiency at the translocated allele at most - and the fusion's phenotype in this graph is a HOXA programme, not an iron-uptake defect. Oncofusion knowledge graph (internal entry) 2026
7. E7strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

   The clathrin assembly protein PICALM is required for erythroid maturation and transferrin internalisation in mice.

   The clathrin assembly protein PICALM is required for erythroid maturation and transferrin internalization in mice. · Suzuki M, et al. · PLoS One 2012
   PMID 22363754DOI
   ▸What this evidence cannot show

   Independent mouse knockout study reaching the same erythroid conclusion as [E6]. Two groups, concordant - which is why this is 'strong' rather than 'moderate'.

   ▸Evidence that argues against this (1)

   • ERYTHROID MATURATION IS THE WRONG LINEAGE FOR THIS GRAPH'S ARGUMENT, and it is the wild-type protein again. PICALM::MLLT10 is commonest in T-ALL and is used here as a KMT2A-free comparator for infant B-lineage disease; an erythroid requirement adds a third lineage to a comparison that already carries one confound. Oncofusion knowledge graph (internal entry) 2026
8. E8strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.review

   The PICALM locus is an established late-onset Alzheimer's disease risk locus, with proposed mechanisms in endocytosis, autophagy and amyloid-beta clearance.

   PICALM and Alzheimer's Disease: An Update and Perspectives. · Ando K, et al. · Cells 2022
   PMID 36552756DOI
   ▸What this evidence cannot show

   This is a REVIEW, cited as a map of a field outside this beat and not as primary evidence. A GWAS association at a locus is not a statement about the protein's mechanism. Included because it bounds the tolerability of any PICALM-directed intervention.

   ▸Evidence that argues against this (1)

   • The Alzheimer's association is a common-variant locus association and the causal gene and direction of effect at the locus are not settled by it; treating it as a statement that lowering PICALM is dangerous, or protective, overreads a GWAS signal. Ando K et al. 2022
9. E9established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

   Every CALM/PICALM structure in the PDB (3ZYK, 3ZYL, 3ZYM, 7JXV, 1HF8, 1HFA, 1HG2, 1HG5) is the RAT protein (UniProt O55012); an RCSB search scoped to the human accession Q13492 returns zero entries.

   RCSB PDB free-text and accession-scoped searches for CALM/PICALM, and entry records 3ZYK, 3ZYM, 1HFA, 7JXV · RCSB PDB 2026
   ▸What this evidence cannot show

   Queried live 2026-09-20. The accession-scoped RCSB query returned HTTP 204 with an empty body, which per the brief's PDB rules is the real answer (zero) rather than a network failure; the free-text query then returned four PICALM entries, all rat, and 3ZYM is additionally a single-chain engineered chimera (mouse VAMP8 residues 11-41 fused to rat CALM 1-264 through an 8-residue linker), so it models a bound state rather than measuring association. Rat and human ANTH domains are close, so the structures are usable - but any figure built from them is a rat structure and must say so.

   ▸Evidence that argues against this (1)

   • EVERY DEPOSITED STRUCTURE IS OF THE SAME MODULE, AND IT IS THE ONE THE FUSION KEEPS LEAST OF. A PDB census showing all CALM/PICALM entries are the ANTH/ENTH-family membrane-binding domain means the structural literature simply does not cover the C-terminal region that PICALM::MLLT10 retains. Structural coverage here is evidence about what crystallises, not about what matters in the fusion. Oncofusion knowledge graph (internal entry) 2026
10. E10strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In murine bone-marrow retroviral transformation assays, CALM residues 544-553 constitute a nuclear export signal that mediates cytoplasmic localisation of CALM-AF10 and is necessary for CALM-AF10-dependent transformation; fusing the CALM NES alone, or heterologous NES motifs from ABL1, Rev, PKIA or APC, to AF10 is sufficient to immortalise murine bone-marrow cells.

    A CALM-derived nuclear export signal is essential for CALM-AF10-mediated leukemogenesis. · Conway AE, et al. · Blood 2013
    PMID 23487024DOI
    ▸What this evidence cannot show

    Murine bone-marrow retroviral transformation plus HEK293 imaging - not a human patient system, and not this project's index infant B-lineage KMT2A::AFF1 disease.

    ▸Evidence that argues against this (1)

    • A NUCLEAR EXPORT SIGNAL DEFINED IN A MURINE RETROVIRAL TRANSFORMATION ASSAY INHERITS THAT ASSAY'S LIMITS. Residues 544-553 are identified by what their mutation does to transformation in overexpressing murine bone marrow, which is a functional definition in one system rather than a demonstration that the endogenous human fusion uses that NES. This graph's open export-versus-docking question turns on exactly this point and is not settled. Oncofusion knowledge graph (internal entry) 2026
11. E11moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset

    In DepMap CRISPR leukaemia screens, PICALM gene effect in the CALM-AF10-positive line U-937 is -0.345, within the spread of non-KMT2A-rearranged leukaemia lines (mean -0.261, SD 0.324, n=42).

    DepMap CRISPR (Chronos) gene effect via the Open Targets Platform GraphQL API · DepMap / Open Targets Platform 2026
    ▸What this evidence cannot show

    n=1 CALM-AF10 line with screen data (U-937; the second Cellosaurus-annotated line, TUR / ACH-002314, has no screen). A gene-level CRISPR score cannot distinguish wild-type PICALM from the 5' portion retained in the fusion, so this is not a test of fusion dependency. The DepMap release version is not stated by the Open Targets API.

    ▸Evidence that argues against this (1)

    • ONE CELL LINE, AND IT IS THE SAME LINE EVERY OTHER CALM-AF10 DEPENDENCY CLAIM HERE RESTS ON. U-937 is n=1 and is histiocytic/myeloid while PICALM::MLLT10 is commonest in T-ALL. This project's own power analysis put the minimum detectable difference above the observed difference for the companion MEN1 value, so a gene-effect number from this line is consistent with a wide range of truths. Oncofusion knowledge graph (internal entry) 2026
12. E_q77_loc1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    PICALM's subcellular record in UniProt Q13492 (fetched 2026-09-21) is FULLY EXPERIMENTALLY CODED - five locations, every one at ECO:0000269 with a PubMed citation, and no ECO:0000250 or ECO:0000305 anywhere in the block: cell membrane; membrane/clathrin-coated pit; Golgi apparatus; cytoplasmic vesicle/clathrin-coated vesicle; and nucleus. The founding cell-biology observation for the first four is Tebar et al. 1999, whose title is itself a localisation claim.

    Clathrin assembly lymphoid myeloid leukemia (CALM) protein: localization in endocytic-coated pits, interactions with clathrin, and the impact of overexpression on clathrin-mediated traffic. · Tebar F, Bohlander SK, Sorkin A · Mol Biol Cell 1999
    PMID 10436022DOI
    ▸What this evidence cannot show

    Rule 91 check performed and PASSED: the cited paper was opened and is about the annotated fact - its title is 'localization in endocytic-coated pits'. That is unusual in this tranche and is recorded as the positive control for the check. Limits of the paper itself: immunofluorescence and overexpression in cultured non-haematopoietic cells, so it establishes compartment rather than physiological requirement and says nothing about haematopoietic cells. The UniProt comment block was read on 2026-09-21; the accession and length were already asserted in this entry and match.

    ▸Evidence that argues against this (1)

    • A FULLY CODED UNIPROT BLOCK STILL DOES NOT MAKE THE ANNOTATION A MEASUREMENT IN YOUR SYSTEM. All five locations come from transfected or depleted cultured cells, none haematopoietic, and the coated-pit result is an over-expression experiment whose own follow-up narrowed the physiological requirement to the erythroid lineage. Well-coded and irrelevant to a blast are not mutually exclusive. Ishikawa Y et al. 2015
13. E_q77_loc2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    The SOLE experimental citation UniProt gives for wild-type PICALM being nuclear is a paper about the FUSION's localisation: Archangelo et al. 2006 identify CATS (FAM64A) as a CALM interactor binding CALM aa 221-335 - a region retained in CALM/AF10 - report that CATS localises to the nucleus with a preference for NUCLEOLI, and show that CATS expression markedly increases the nuclear localisation of CALM AND of CALM/AF10.

    The novel CALM interactor CATS influences the subcellular localization of the leukemogenic fusion protein CALM/AF10. · Archangelo LF, Gläsner J, Krause A, Bohlander SK · Oncogene 2006
    PMID 16491119DOI
    ▸What this evidence cannot show

    Rule 91, firing in an unusual direction. The evidence code is correct and the paper IS a localisation study, but its SUBJECT is the fusion, and the wild-type nuclear call is a by-product of an experiment designed to move the fusion. The nuclear pool is also demonstrated under CATS OVER-EXPRESSION, which is a perturbation and not a baseline: the paper does not establish what fraction of endogenous PICALM is nuclear in an unperturbed cell. Yeast two-hybrid, pull-down, co-IP and co-localisation in transfected cells; no primary haematopoietic material. The nucleolar preference belongs to CATS and is NOT shown for PICALM.

    ▸Evidence that argues against this (1)

    • This graph separately records CALM/AF10 as PREDOMINANTLY CYTOPLASMIC, which is the opposite emphasis to a paper reporting that a partner drives it into the nucleus. The two are reconcilable - cytoplasmic at steady state, shifted nuclear by an over-expressed interactor - but only if the reader is told both, and an entry quoting either alone gives a confident wrong impression of where this fusion is. The independent reason to weight the cytoplasmic reading at baseline is the NES result: CALM residues 544-553 are a CRM1-recognised export signal whose loss abolishes transformation. Conway AE et al. 2013
14. E_q77_loc3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    The cell-membrane annotation rests on a DEPLETION study rather than an imaging study: Meyerholz et al. 2005 deplete CALM by RNA interference and read out clathrin coat formation, and UniProt cites it as the ECO:0000269 source for 'Cell membrane'.

    Effect of clathrin assembly lymphoid myeloid leukemia protein depletion on clathrin coat formation. · Meyerholz A, Hinrichsen L, Groos S, Esk PC, Brandes G, Ungewickell EJ · Traffic 2005
    PMID 16262731DOI
    ▸What this evidence cannot show

    Rule 91 partial fail, recorded rather than repaired. The paper is squarely in the right biology, but its design is loss-of-function with a coat-formation readout, so it supports membrane ASSOCIATION by inference from a functional perturbation rather than by direct localisation of the protein. The direct membrane and coated-pit imaging is in Tebar 1999 [E_q77_loc1], which UniProt also cites, so nothing is lost - the point is that an evidence code does not tell you which of the two designs you are getting.

    ▸Evidence that argues against this (1)

    • A depletion phenotype at a compartment is not a statement that the protein is at that compartment; the defect could be indirect. Here the direct imaging exists and agrees, so the objection is about method transparency rather than about the fact. It is recorded because the same generous reading is NOT safe for the five proteins in this tranche whose UniProt blocks carry no experimental code at all. UniProt Consortium 2026
15. E_q77_ct1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    In HUMAN LEUKAEMIA CELL LINES and primary murine bone marrow, CATS/FAM64A - the interactor that drives PICALM and CALM/AF10 into the nucleus - is differentiation-coupled and clonogenicity-limiting: its expression falls during erythroid, megakaryocytic and monocytic differentiation and rises during ATRA-induced granulocytic differentiation; lentiviral knockdown in U937 reduces proliferation, alters cell-cycle progression, lowers migration and REDUCES CLONOGENICITY with reduced GLI-1; and retroviral over-expression of murine Cats in primary bone marrow ALSO decreases colony formation.

    CATS (FAM64A) abnormal expression reduces clonogenicity of hematopoietic cells. · Barbutti I, Xavier-Ferrucio JM, Machado-Neto JA, Ricon L, Traina F, Bohlander SK, Saad ST, Archangelo LF · Oncotarget 2016
    PMID 27588395DOI
    ▸What this evidence cannot show

    Knockdown and over-expression both reduce clonogenicity, which is non-monotonic and weakens any directional reading; the authors' own xenograft arm was negative, so the in vitro effects did not translate. U937 is the same single line that every CALM-AF10 dependency claim in this graph rests on, and it is histiocytic/myeloid while PICALM::MLLT10 is commonest in T-lineage disease. This is cell-type evidence about CATS, imported here because CATS is the mechanism of PICALM's nuclear pool; it is not evidence about PICALM's own requirement.

    ▸Evidence that argues against this (1)

    • The physiological cell-type requirement for PICALM itself points somewhere else entirely. Two independent murine knockouts confine it to the ERYTHROID lineage - transferrin-receptor endocytosis and erythroid maturation - with myeloid and B-lymphoid development intact. The gene's own genetics says erythroid, the interactor's cell biology says myeloid lines, and neither says anything about the T-lineage compartment where PICALM::MLLT10 disease actually arises. Ishikawa Y et al. 2015

• ← part ofPICALM / CALM - domain architecture, residue coordinates, and which parts the fusion keeps

  domain map of UniProt Q13492 (652 aa)

• fuses with →MLLT10 (AF10) — DOT1L cofactor and KMT2A fusion partner

  t(10;11)(p13;q14) joins PICALM 1-648 to the 3' portion of MLLT10

• ← part ofPICALM::MLLT10 (CALM-AF10), t(10;11)(p13;q14) - the DOT1L-dependent HOXA fusion that contains no KMT2A

  5' partner, residues 1-648

• ← part ofA ten-residue export signal is the 5' partner's whole contribution - and the field disagrees about which direction it pushes

  NES at Q13492-1 residues 544-553

• ← part ofWhere wild-type PICALM actually is: coated pits, the Golgi, coated vesicles - and a nuclear pool nobody has explained

  where the wild-type protein is

• ← related toThe non-chromatin half: CALM-AF10 breaks the cell's iron intake and the cells are chelator-sensitive in vitro - but iron restriction did not extend survival in vivo (corrected 2026-09-21)

  the wild-type function the fusion fails to perform, and acts dominant-negatively against

• ← related toCALM-AF10 is predominantly cytoplasmic in transfected non-haematopoietic cells - and NUCLEAR in leukaemia cells (title corrected 2026-09-21; the five systems are five instances of one design)

  the CALM aa 221-335 CATS-binding region is retained in the fusion and CATS expression drives the fusion nuclear, so cell-line CATS level is an uncontrolled variable in every image

• ← related toAP2B1 (beta-2-adaptin) - the only one of AFF4's four two-hybrid partners that is not a hub, and the only one with an echo in this graph's endocytosis biology

  AP-2 is the clathrin adaptor PICALM works alongside; recorded as a pathway-name coincidence and rejected, not proposed

• related to →Where wild-type PICALM actually is: coated pits, the Golgi, coated vesicles - and a nuclear pool nobody has explained

  PICALM's own subcellular record: fully ECO:0000269-coded across five compartments, nuclear pool included

• related to →CALM-AF10 is predominantly cytoplasmic in transfected non-haematopoietic cells - and NUCLEAR in leukaemia cells (title corrected 2026-09-21; the five systems are five instances of one design)

  the fusion's compartment, which wild-type PICALM's cytoplasmic baseline predicts and the CATS result complicates

• related to →A partner's 'normal' compartment is usually read off a different tissue, and for at least two KMT2A partners that reading is wrong in blood cells

  PICALM is the clean positive instance of the warning: an endocytic, cytoplasmic partner whose fusion acts on chromatin

• expressed in →Erythroid progenitor - the committed red-cell precursor, distinct from the bipotent MEP that two entries had to borrow, and the compartment that pays for any haem-pathway or fetal-globin intervention

  murine conditional knockout: Picalm is required for transferrin-receptor endocytosis and erythroid maturation, and dispensable for myeloid and B-lymphoid development

• related to →The cell CALM-AF10 arrests: an early T-precursor in adults and a TCR-positive thymocyte in children - two states, split by age (corrected 2026-09-21)

  the state belongs to the FUSION, not to PICALM - any heterologous NES substitutes for the CALM contribution